Hierarchical Transformer for Brain Computer Interface

In this paper, we propose a novel hierarchical trans-former classification algorithm for the brain computer interface (BCI) using a motor imagery (MI) electroencephalogram (EEG) signal. The reason of using the transformer-based is catch the information within a long MI trial spanning a few seconds, and give more attention to the time periods during which the intended motor task is imagined by the subject without any artifact. The hierarchical transformer architecture consists of a high-level transformer (HLT) and a low-level transformer (LLT). We break down a long MI trial into a number of short-term intervals. The LLT extracts a feature from each short-term interval, and the HLT pays more attention to the features from more relevant short-term intervals by using the self-attention mechanism of the transformer. We have done extensive tests of the proposed scheme on two open MI datasets, and shown that the proposed hierarchical transformer achieves outstanding results.